In Situ Visualization of Lithium Ion Intercalation into ... - ACS Publications

Subscriber access provided by KUNGL TEKNISKA HOGSKOLAN

Article

In situ Visualization of Lithium Ion Intercalation into MoS2 Single Crystals using Differential Optical Microscopy with Atomic Layer Resolution Azhagurajan Mukkannan, Tetsuya Kajita, Takashi Itoh, Youn-Geun Kim, and Kingo Itaya J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b11849 • Publication Date (Web): 16 Feb 2016 Downloaded from http://pubs.acs.org on February 18, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

In situ Visualization of Lithium Ion Intercalation into MoS2 Single Crystals using Differential Optical Microscopy with Atomic Layer Resolution Mukkannan Azhagurajan,† Tetsuya Kajita,‡ Takashi Itoh,‡ Youn-Geun Kim,§ and Kingo Itaya*,†,‡ †

Institute of Multidisciplinary research for Advanced Materials, Tohoku University, Katahira Campus, Sendai 980-8577, Japan. ‡

Frontier Research Institute for Interdisciplinary Sciences (FRIS), Tohoku University, 6-3 Aoba, Sendai 980-8578, Japan

§

Joint Center for Artificial Photosynthesis California Institute of Technology, Division of Chemistry and Chemical Engineering, Pasadena, CA 91125, United States.  Supporting Information ABSTRACT: Atomic-level visualization of the intercalation of layered materials, such as metal chalcogenides, is of paramount importance in the development of high-performance batteries. In situ images of the dynamic intercalation of Li ions into MoS 2 singlecrystal electrodes were acquired for the first time, under potential control, with the use of laser confocal microscopy combined with differential interference microscopy technique (LCM-DIM). Intercalation proceeded via a distinct phase separation of lithiated and de-lithiated regions. The process started at the atomic steps of the first layer beneath the selvedge and progressed in a layer-by-layer fashion. The intercalated regions consisted of Li-ion channels into which the newly inserted Li ions were pushed atom-by-atom. Interlayer diffusion of Li ions was not observed. De-intercalation was also clearly imaged and was found to transpire in a layer-bylayer mode. The intercalation/de-intercalation processes were chemically reversible and can be repeated many times within a few atomic layers. Extensive intercalation of Li ions disrupted the atomically flat surface of MoS 2 due to the formation of small lithiated domains that peeled off from the surface of the crystal. The current-potential curves of the intercalation/de-intercalation processes were independent of the scan rate, thereby suggesting that the rate-determining step was not governed by Butler-Volmer kinetics.

 INTRODUCTION

electrical properties of Li interaction into MoS2 nano-sheets was also investigated.15,16

Direct visualization of processes on the electrode surface provides a crucial guide in the rational control and optimization of charge-transfer events. In situ scanning probe microscopic techniques such as scanning tunneling microscopy (STM) and atomic force microscopy (AFM) are well-established techniques that impart atomic-level information of various interfacial phenomena.1-5 Experimental techniques based on Raman, infrared, and X-ray scattering spectroscopies have been specifically modified also to track the dynamic processes of electrochemical reactions.6 STM and AFM studies have been previously applied to the intercalation/de-intercalation reactions of Li-ions and anions into highly ordered pyrolytic graphite.7-9 Electrochemical deposition of copper and organic molecules into transition metal dichalcogenides such as TaS2 has also been investigated which inferred the formation of nanoscale defects during organic molecule intercalation/deintercalation.10,11 MoS2 has been previously studied and employed in many applications, such as catalysis, batteries and solid lubricants.12,13 Recently, MoS2 has attracted much attention in energy storage applications due to its specific physico-chemical properties.13,14 The optical and

However, a handful of practical limitations of STM and AFM often curtails the wide applicability of these techniques: (i) The acquisition times, typically in the order of minutes, are known to be inadequate in the evaluation of fast, dynamic processes. Note that a video-rate in situ STM technique has been developed recently to overcome this problem.17,18 (ii) The small observable scan areas, usually at sub-micrometer scale, necessitate the burden of proof that the acquired images statistically represent the entire electrode surface. (iii) STM tips and AFM cantilevers may perturb the concentration distribution of the solute in the vicinity of the electrode surface, thereby leading to, under certain experimental conditions, sluggish electrochemical deposition of metals19,20 and unfavorable effects during the crystal growth of organic materials.21 We have recently developed a laser confocal microscope combined with a differential interference contrast microscope (LCM-DIM) that can resolve single atomic steps with ca. 0.20.3 nm step heights on commonly used metal electrodes. With acquisition times of 2–10 frames per second, the LCM-DIM was shown to capture images of the dynamic electrochemical

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

dissolution and deposition processes on Au(111) surfaces as described in our previous papers.22,23 Atomically flat surfaces that extended more than a few micrometers were needed to demonstrate clearly the capabilities of the method, because the inplane resolution resembled that of normal optical microscopes at ca. 0.2–0.3µm.22 In this regard, protocols for the preparation of ultra-flat Au, Pd, Pt, and Si(100) electrodes were developed. 22-26

The intercalation of layered materials, such as MoS2, for charge-storage purposes is an archetypal process that benefits from the direct visualization by LCM-DIM. Efforts to improve the performance of lithium-ion batteries conventionally evaluate current-voltage curves and impedance measurements 12-14 which are devoid of any direct structural information. The elechrochemical STM (EC-STM) has already uncovered important atomic details of the initial stages of lithium-ion intercalation,79 but a complete visualization of the intercalation/de-intercalation cycle has never been accomplished. The present report showcases a gallery of LCM-DIM images and videos that reveal the progression of Li-ion intercalation and de-intercalation initiated at the monoatomic steps of MoS2 in an ionic liquid. The acquisition of such images with atomic-height resolution has unveiled mechanistic insights of the process.  MATERIALS AND METHODS The LCM-DIM used herein was an improved version of the configuration first reported by Sazaki.21,27 Details of the set-up are described in Figures S1 and S2 of the Supporting Information. The acquisition time of each image was typically less than a second, depending on the number of data points. The LCM-DIM optics was basically constructed for measurements in air. To achieve atomic-height resolution in solution, a specially designed objective lens (LUCPLAN FLN, Olympus) with a compensator was necessary to account for changes in the refraction index at the solution-glass interface. A high-resolution image of the electrode surface can, therefore, be obtained through a relatively thick layer (< 2 mm) formed by the solution and the glass base plate of the electrochemical cell. A more detailed description of the experimental set-up is found in Figure S2. AFM images were acquired using a Pico-SPM 5500 (Agilent Corp.). High-resolution STM images were obtained by a NanoScope E (Digital Instruments, Santa Barbara, CA). MoS2 has three well-known polytypes: 1T-MoS2, 2H-MoS2 and 3R-MoS2. X-ray powder diffraction indicated that the MoS2 sample under investigation had the 2H-MoS2 structure. The MoS2 crystals were readily cleaved to expose ultra-flat surfaces with terraces wide enough to observe atomic steps clearly with LCM-DIM. The lateral (in-plane) resolution of LCM-DIM was in the range of 0.2–0.3 μm, akin to conventional optical microscopes. Current-potential profiles were acquired using a BAS potentiostat 700D. Details of the electrochemical cell are displayed in Figures S1-S3. An O-ring (Kalrez, Dupon), with outer and inner diameters of 8 mm and 5 mm, respectively, was placed between the MoS2 crystal and the electrochemical cell to create a vacuum-tight seal. Li metal was used as a reference electrode. The electrolyte solution consisted of 0.32 mol lithium bis(trifluoromethylsulfonyl)imide (LiTFSI) (Kanto Chem. Co.) per kg of 1-ethyl-3-methylimidazolium bis(fluorosulfonyl)imide

Page 2 of 9

(EMI-FSI). To ensure the removal of water, the ionic liquid solution was heated inside the vacuum-tight electrochemical cell at 130°C using an oil bath for more than 10 hours until the pressure was lower than 1 x 10-5 Pascal. A turbo-molecular pump (PFEIFFER Mod. HiCube 80 Classic) was employed to keep the base pressure lower than 10-5 Pascal. For LCM-DIM measurements, Au and Teflon meshes were placed between the optically flat glass plate and MoS2. Details of this configuration are described in Figure S3. For reference purposes, the voltammetric behavior (Figure S4b) of MoS2 was acquired in a batterygrade electrolyte solution of 1:1 (by volume) ethylene carbonate and diethyl carbonate (Kanato Chem. Co.) containing 1 M LiClO4, without further purification, because most of the previous works on Li-ion batteries used this solvent system.12,13

 RESULTS Cyclic voltammetry of basal planes. As-received ionic liquids were found to produce a large cathodic current prior to the onset potential for Li-ion intercalation. For instance, the background current at the potential range between 1.4 V and 2.8 V reached ca. 50 μA cm-2 at 10 mVs-1 (red trace of Figure S4a). Such large background currents were strongly dependent on the concentration of water; preliminary experiments inside a glove box revealed that the ionic liquid solution contained 20-50 ppm water. The use of rigorously dried solvents dramatically decreased these pre-intercalation peaks by more than two orders of magnitude. Figure S4a indicates that the persistence of broad cathodic peaks should be due to trace amounts of water. A similar small background current has been observed on a highly ordered pyrolytic graphite (HOPG) in the same ionic liquid.28 In a rigorously dried solvent, the cathodic current for intercalation commenced at potential of 1.3 V vs. Li, and increased as the electrode potentials were swept in the negative direction from 3.2 V vs. Li, as shown in Figure 1a. De-intercalation was induced in the reverse potential scan, giving rise to an anodic peak at ca. 1.6 V. These experiments demonstrated the chemical reversibility of the intercalation and de-intercalation processes. Surprisingly, both anodic and cathodic currents did not vary significantly with the scan rates between 10 to 100 mVs-1. The diffusion of Li ions into the MoS2 sheets was, therefore, nonFickian, otherwise the currents would have been proportional to the square root of the scan rate. At the lowest scan rate of 10 mVs-1 (black trace in Figure 1a), a small cathodic current appeared at ca. 2.2 V. At a potential range of 1.1 V - 1.3 V, the intercalation current was almost linearly dependent on the potential. The Butler-Volmer equation predicts an exponential dependence of current on potential for electrode processes controlled by electronic charge transfer. Figure 1b depicts the relationship between the total charges associated with the intercalation and de-intercalation processes. Line (A) represents measured charges without background correction. According to this line, only ca. 80% of doped Li ions were de-intercalated. A close scrutiny of Figure 1a indicated small cathodic currents (< 1 μA cm-2) between the onset potential for intercalation and 2.2 V; such small currents were ascribed to the reduction of trace amounts of water or unknown organic impurities.


Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society each STM image showed in figure 3a’ and 3b’. Figure 3a shows an atomically flat terrace with a step height of ca. 0.62 nm, which was consistent with the literature value of the height of the unit cell in the c-axis of 2H-MoS2.12,13

Figure 1. (a) Cyclic voltammograms of a freshly cleaved MoS2 in a rigorously dried ionic solvent at different scan rates. The electrode potential was scanned from 3.2 V vs. Li. (b) Relationship between intercalation and de-intercalation charges. Red line (A) shows the raw data without the correction for the small currents between 2.2 V and 1.45 V. Black line (B) represents the corrected data.

Control experiments were conducted in LiTFSI-free solvents to ascertain background-charge contributions. Background-correct charges (Line (B)) revealed that nearly all of the intercalated Li ions were involved during the charging-discharging processes. For 2H-MoS2 crystal, the maximum charge density for a single monolayer is 0.185 mC cm-2, calculated from the structure of 2H-MoS2. This value corresponds to the theoretical specific charge capacity of 167 mAh g-1.12,29 The highest charge density measured in the experiments shown in Figure 1b was 2 mC cm-2, which was obtained at a scan rate of 10 mVs -1; this charge corresponded to the intercalation of ca.10 monolayers. LCM-DIM images. MoS2 crystals can be cleaved easily by an adhesion tape (such as Scotch tape). Surfaces cleaved in this manner usually showed many step lines and atomically flat regions with narrow widths. Ultra-flat wide surfaces, however, can be exposed by the placement of the crystal between, and its subsequent detachment from, two glass plates each stuck with an instant adhesive. LCM-DIM images of perfectly cleaved MoS2 samples (Figure S5) clearly show well-defined ultra-flat surfaces with only monoatomic steps. Figures 2a and 2b show two typical wide-scan images (140 µm x 140 μm) acquired by LCM-DIM in air. The surfaces were remarkably flat, and the terraces were more than several micrometers wide. Notably, the MoS2 sample was polycrystalline, but single-crystal domains were obviously large. All step lines formed during cleavage were random and showed no preferential direction with respect to the crystal structure. Large pits surrounded by monoatomic steps, such as those bordered by the dashed circle S1 in Figure 2a, originated from wide terraces that were partially peeled off during the cleaving process. Monoatomic steps typically registered an atomic height of 0.6 nm. Some monoatomic steps (S1 in Figure 2b) were found to extend more than 100 µm across the surface. Diatomic (S2) and multi-atomic steps (S3 and S5) were also found in different areas. It is noteworthy that when the step height was smaller than the wavelength of the laser, the contrast (brightness) of these steps almost scaled linearly with the step height. Such linearity was confirmed by the comparison between LCM-DIM and STM/AFM images. STM images. Verification of the step heights observed by LCM-DIM was accomplished by the re-examination of the same samples with EC-STM operated in air. Figure 3 displays two typical STM images. Height profiles along lines drawn in

Figure 2. Typical LCM-DIM images of freshly cleaved surfaces observed in air. The size of observed areas were 140 x 140 µm. Both images show very wide terraces with mostly monatomic steps (S1); steps with multi-atomic heights, such as S5, can also be found. (a) Basin-shaped pits surrounded by monoatomic steps are indicated by the dashed circle, S1. (b) Two monoatomic step lines marked by S1 are relatively straight and extended more than 100 µm across the surface. All images were taken in a large area of 140 µm x 140 µm.

Figure 3b displays a surface with several steps of different atomic heights. The S1 and S2 arrows indicated monoatomic and diatomic steps, respectively. The coalescence of two diatomic steps gave rise to a step (S4) that is four-atoms high.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 9

It has been reported that MoS2 nanosheets with different thicknesses showed decrease in sheet resistances.16 Therefore, it is expected that the first layer might be intercalated at the first stage. Simultaneous intercalation at the underlying layers would have produced domains of varying step heights. Furthermore, these results negate the possibility of Li-ion diffusion into the bulk of the material. If Fick’s law of diffusion occurred across intercalated domains, the domain boundaries would have shown a gradual decrease in contrast. The acquired images indicated that the lithiated and de-lithiated domains led to the formation of two distinct phases. A similar phase separation was previously observed in the doping of Li with TiO2.30

Figure 3. High-resolution STM image of MoS2 in ionic liquid at a potential of 3.0 V. The tunneling current was 5 nA. (a) The height profile along a line S1-S1’ indicated that the step is monatomic with height of ca. 0.62 nm as shown in (a’). (b) The height profile along a line S2-S2’ revealed the presence of diatomic steps as shown in (b’). These diatomic steps coalesced at point S4 to form a step that was four-atoms high. The images of (a) and (b) were taken in an area of 10 µm x 10 µm and 7 µm x 7 µm, respectively.

A careful inspection of the high-resolution STM images of the terraces in Figure 3b revealed regular lines with height differences (ca. 0.1 - 0.2 nm) smaller than a monoatomic step height (ca. 0.62 nm). Details from Figure S6a and S6b showed that these line features constituted steps with sub-monoatomic heights. Intercalation and de-intercalation in the first monolayer underneath the selvedge. A slow potential sweep (scan rate = 10 mV s-1) headed in the negative direction from 3.2V vs. Li immediately gave rise to a cathodic current at 1.3V that marked the intercalation of Li ions into the MoS2 crystal. At the incipient stage of intercalation, when the current density was only ca. 1-2 μA cm-2(1.25 V vs. Li), the electrode potential was fixed and a series of LCM-DIM images (Figures 4(a-c)) was acquired. Note that in order to see two phases clearly, a data treatment was applied with high contrast ratio. In this case, terraces seemed to be not very flat, shows rolling hill structures. This structure might be an artifact caused by the present data treatment. A similar irregular structure can also be seen in Figure 2a. A new domain with a dark contrast appeared at the upper left side of Figure 4a and expanded, in the direction indicated by the arrow, into the inner part of the terrace. Another domain with the same dark contrast developed at the lower step (Figure 4b). The dark contrast ultimately became uniform throughout the terrace of interest (Figure 4c). Video S1 in the Supplementary Materials captures the dynamic process in real time. Note that the yellow arrows showed the relatively faster growing direction only. The new domains were also expanding in other directions. Because the step edges were not always straight which might have included kink sites, the intercalation reaction seems to be strongly depends on the densities of kink sites at step edges. It is also expected that the intercalation reaction might be faster at kink sites than those at the straight steps. A further detailed analysis of growth rates is still under investigation. The absence of variations in the image contrast within the intercalated region strongly suggests that the intercalation initially proceeds within the first layer below the selvedge.

Figure 4. Dynamic processes of intercalation (a-c) and de-intercalation (d-f) of the first intercalated layer of MoS2. Each image was acquired at a rate of 1 frame/s (data points: 512 x 512). The potential was swept from 3.2 V at a scan rate of 10 mVs-1. The time indicated in each image started from the appearance of the new domain. The images of (a-c) were obtained at 1.25V vs. Li. De-intercalation (d-f) was observed at 1.35 V, when the current was also ca. 1-2 μA cm-2. All images were taken in an area of 70 µm x 70 μm.

Previous experiments with LCM-DIM revealed that two adjacent terraces bordered by a step would appear as regions with the same intensity demarcated only by step lines; the sharp contrast at the step was an aftermath of the phase shift of two beams produced by the Nomarski prism at the step.27 For this reason, LCM-DIM has been used for the detection of steps on the surface of a material where the upper and lower terraces have the same physical property. Monoatomic step heights, as small as 1.4 Å like in the case of Si(100), can be detected by LCM-DIM 25 . Thus, the interlayer distances of the individual sheets in MoS2 (0.62nm) and its lithiated form, LiMoS2 (0.629nm), might be discernable by present optical technique. 13 However, the images shown in Figure 4 did not showed only steps as observed on Au(111)(described in previous works).22,23 It is clear that new domains with different contrasts appeared during the intercalation reaction. MoS2 shows absorption bands in the visible region.15,16 Such absorption peaks are changed by the intercalation of Li ions.12,31-33 It is reasonably expected that the appearance of different contrast in the intercalated regions are also caused by the difference of the adsorption spectra. To the best of our knowledge, the present study is the first demonstration of a further ability of LCM-DIM to monitor contrast differences on the surface domains which might have caused by the change in optical properties. Such significant


Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


demonstration creates further interest to the researchers to investigate the in situ surface reactions caused by changes in optical properties in the monolayer regime. In addition to the present work, our on-going surface investigation further focus on the visualization of redox reactions in several organic monolayers. After the completion of the intercalation in the first layer below the selvedge, the potential was scanned in the positive direction and was set at 1.35 V, when the current density was only ca. 1-2 μA cm-2. Figures 4(d-f) shows the restoration of the image contrast on the terrace as the Li ions were de-intercalated. The rate of disappearance of the Li-intercalated domains was noticeably not uniform. A careful inspection of Video S1 reveals that the intercalation from the upper left corner to the lower right corner seemed faster than other directions. The de-intercalation reaction was also found to occur with different rates. This behavior can be ascribed to the direction along the b-axis in MoS2. The arrows in Figure 4 indicates the directions where the rate of the intercalation and de-intercalation reactions were relatively faster than those in other directions. However, the presence of kink sites made the steps to deviate from the straight path. Further work on the evaluation of the rate of reactions is still needed to understand this uniaxial process along the channels. Significantly, the rate of reactions depend on the electrode potential as expected from Figure 1a. Note that terraces shown in Figures 4 and 5 there were local variations in contrast. However, these contrast were showed clearly by applying a high contrast in the data treatment to make clear contrast difference between the doped and un-doped areas. Therefore the local variation in contrast seems to be artificial. If there were some defects, the reactions might have also started on terraces. Such modulations can be also seen in Figure 2a. Intercalation and de-intercalation of multilayers. After acquisition of the images shown in Figure 4, the electrode potential was swept in the negative direction (only 2-3 mV) to induce further intercalation. Figures 5(a-c) depict the progressive intercalation of multilayers. Domain l1 in Figure 5a represented the selvedge under which lay the first intercalated layer. The involvement of n underlying layers (where n = 2, 3, 4) triggered the emergence of other domains (ln) with distinct image contrasts. Figure 5 is one of the example of the intercalation reactions. We expected that many steps might be existed at near the left lower corner or out of the observed area, which became reaction sites to form multilayers. In addition, along the axes of a and b shown in Figure 6 have the equivalent crystal structure, which caused the 2D growth of domains. These multilayers seemed to initiate from lower right corners where we believe that reactions were also started from step edges located at these corners. We added one more example of videos (S5), in which multilayer formations occurred in different directions of the intercalation. The corresponding de-intercalation process was monitored in Figures 5(d-f). Domain l4 disappeared first; domains l3 and l2.subsequently followed. The de-intercalation process occurred layer-by-layer, ultimately creating an atomically flat surface with Li ions still intercalated in the first monolayer underneath the selvedge. At more positive potentials, exhaustive de-intercalation of the remnant layer transpired akin to the process captured in Figures 4(d-f).

These processes are fully shown in Movie S2 in the Supporting Information. Figure 6 illustrates a proposed mechanism of the intercalation and de-intercalation processes of Li ions in MoS2 based on the images provided by in situ LCM-DIM.

Figure 5. Dynamic processes of the intercalation and de-intercalation reactions involving multilayers. After the intercalation of the first layer as shown in Figure 4c, the electrode potential was shifted to the anodic direction by only a few mV (2-3 mV). The second, third, and fourth layers marked by l2, l3, l4 started to be intercalated in a layer-by-layer mode (a-c). At more positive potentials (1.40 V), exhaustive de-intercalation of the remnant layer transpired akin to the process captured in (d-f). These layers exhibited different image contrasts, thereby allowing a distinct delineation of the intercalated and de-intercalated domains (or phases). All images were taken in an area of 70 µm x 70 µm.

To save clarity of the mode, we drew that the step edge is straight without kink sites. And, the structural model in Figure 6 shown along the b axis. It is located along the a and b axes at 120° angles to each other, the structure along the b axis should be the same as a. Therefore, the growth of domains should be occurred in two dimensional diffusion. Moreover, the edges should have irregular shape, including many kink sites. These kink sites are expected to cause quasi-two dimensional growth of domains as described above. (i) Li ions are initially incorporated at the steps of the MoS2 crystal. (ii) Intercalated Li ions accumulate to form of a monoatomic row along the b axis that moves toward the interior portion of the MoS2 sheet. (iii) At the early stage of intercalation, Li ions occupy all available sites along the b axis; i.e. they do not readily diffuse to other sublayers. In effect, intercalated domains (or phases) are distinctly delineated from those that remain intercalant-free. (iv) Current densities for both intercalation and de-intercalation almost linearly increased with potential. This suggests that the rate-determining process is not governed by Butler-Volmer kinetics but is an interplay of two factors: the charge transfer process at the step and the force that pushes a row of Li ions into the interior of the basal plane of the crystal. Extensive Intercalation Processes. Thus far, the LCM-DIM images were acquired only at very low current densities (1-2 µA cm-2) that induce intercalation and de-intercalation within only a few monolayers from the MoS2 selvedge. To reveal surface structural changes associated with high concentration of intercalated ions, the electrode potential was scanned until the peak current for intercalation reached ca. 12 µA cm-2. The black trace in Figure 7a was obtained when the potential was scanned up to 1.0 V from the 2.5 V; included for comparison was the voltammogram in red trace similar to the one shown in Figure 1a. The



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

cathodic peak at ca. 1.1 V was the signpost for intercalation; the anodic peak at 1.45 V marked the de-intercalation process.

Page 6 of 9

area (Figure 8a) rapidly developed reticulated (net-like) features with variant contrasts, indicative of fast intercalation that involved underlying multilayers. The reticulated domains spread throughout the surface as can be seen in Figures 8(b-c). Despite the reticulation, the intercalated layers appeared as dark domains, in a manner similar to what was observed at low intercalation current densities. The total amount of charge was 3 mC cm-2 at the stage shown in Figure 8c.

Figure 6 An illustrative depiction of the Li-ion intercalation of MoS2 along the b axis. The first layer underneath the selvedge is intercalated first to (near) completion prior to the intercalation of the second layer (a and b). A solvated Li ion is initially intercalated at the steps. All intercalated Li ions along the b-axis are pushed toward the interior of the crystal. In this example, a row of four intercalated Li ions takes in an incoming Li ion making the raw with five ions.

Figure 7b shows the voltammetric behavior of MoS2 numerously cycled between 1.9 V and 0.9 V. Coulometric analysis indicated that ca. 100 - 150 layers were involved. Currents for both intercalation and de-intercalation increased with the number of potential cycles. The peak potentials for de-intercalation progressively shifted toward the positive direction. It is anticipated that the de-intercalated Li ions in the deeper part of the crystal need larger overpotentials. The concomitant increase in the de-intercalation current was ascribed to the proliferation of step sites that disrupted the atomically flat surface. Topographic details of this surface disruption are evident in the LCM-DIM images of Figure 8.

Figure 7. Current-potential profiles of MoS2 extensively intercalated with Li ions. (a) The red trace shows the cyclic voltammogram of an initial stage of the interaction reaction of ca. four layers. The black trace represents the cyclic voltammogram of a sample containing ca. 20 intercalated layers (total charge: ca. 4 mC cm-2) prepared expanding the potential window to 0.95V. The peaks at 1.1 V and 1.4 V correspond to the intercalation and de-intercalation reactions, respectively. Scan rate for both traces is 50 mVs-1. (b) Repeated voltammetric cycles between 0.9 V and 1.9 V show that the currents progressively increased with increasing number of potential cycles.

The initial surface in Figure 8a was strewn with multiple steps of different heights. The steps marked by S1 were monoatomic. A multi-atomic step was denoted by S4. The boxed-in

The growth of fronts of the smooth multilayer and the net like structures as nearly similar speed. However, a careful inspection indicates that the local growth rates were different. It was also interestingly found further formation of net-like structures observed more clearly. Similar net-like features have been observed on layered surfaces during metal deposition and intercalation in UHV. 10,11,34-36 This behavior has been interpreted by a lattice expansion of the surface layers. 34-36 We are now studying the net-like structures more carefully in electrochemical conditions by changing the electrode potential. The bright veins of the reticulation suggested that Li ions existed in multilayers along these lines. After the acquisition of the image in Figure 8c, the electrode potential was scanned to anodic direction at 50 mV s-1 to spur de-intercalation. Many of the intercalated domains gradually disappeared with concomitant changes in contrast similar to those of Figure 4. However, the reticulation persisted even after 15 min at 1.8 V as shown in Figure 8d. Video S3 in the Supporting Information shows these processes.

Figure 8. Surface evolution of MoS2 as the potential was swept at a scan rate of 10 mV s-1. (a) Initial state at 2.5 V. (b) At 1.1 V, which is the intercalation peak potential shown in the black trace of Figure7a. (c) At 1.0 V, reticulated (net-like) features covered the surface. The de-intercalation (d) was observed, reticulation persisted can be seen even at 1.8 V. All image were taken in an area of 140 µm x 140 µm.

Finally, the electrode potential was stepped from 1.7 V to 0.9 V. The intercalation reaction occurred at a rapid rate. Video S4


Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


shows an example acquired during such fast interaction reaction. Each image was acquired 0.25 sec/frame (data points: 256 x 256). It is clear that such experiments demonstrated that our LCM-DIM can follow the fast electrochemical reactions. Note that the current at 0.9V was ca. 10µA cm-2 which was almost constant during 10 min. The growth of fronts of the smooth multilayer and the net like structures were also seen as shown in Figure 8. However, the rate of the expansion of intercalated smooth domains was roughly 10-15 times larger than that found in Figure 4, because the current at 0.9V is larger than that in Figure 4. The reticulated domains presumably started at defect sites and also spread rapidly. Note that the shape of the reticulated domains seems to be different from those found in Figure 8, suggesting that the lattice expansion depends on the electrode potential. More detailed study is now under investigation.

 ACKNOWLEDGMENT

Finally, it is noteworthy that the interaction and de-intercalation reactions could be imaged for graphite materials which suggests that LCM-DIM is a general method to many battery electrodes.

97, 1129.

 CONCLUSIONS In situ images of the dynamic intercalation of Li ions into MoS2 single-crystal electrodes were acquired for the first time, under potential control, with the use of laser confocal microscopy combined with differential interference microscopy technique (LCM-DIM). Analysis of variations in the image contrast provided a visual indicator that differentiated the intercalated domains from the intercalant-free regions. Li ions initially occupied the steps and proceeded to form channels within the MoS 2 crystal in a layer-by-layer fashion. Within an appreciable potential window between 0.9 V and 1.8 V, the rate-determining step of the intercalation and de-intercalation processes was not governed by Butler-Volmer kinetics. Newly inserted Li ions were pushed atom-by-atom into the channels to form wide domains. The intercalation/de-intercalation processes were chemically reversible and could be repeated many times within a few atomic layers. However, extensive intercalation of Li ions led to the disruption of the atomically flat surface as evidenced by reticulated surface features that developed into cracks that ultimately peeled off from the surface. The demonstrated capability of LCM-DIM ushers in new possibilities for the direct observation of surface reactions accompanied by changes in optical properties in the monolayer regime.

 ASSOCIATED CONTENT  Supporting Information. The Supporting Information is available free of charge on the ACS Publications website at DOI: xxxxxxxxxxxxx. Experimental Procedure and additional experimental results

 AUTHOR INFORMATION Corresponding Author *[email protected]

The authors acknowledge Prof. G. Sazaki (Hokkaido University), Mr. Y. Saito (Olympus), Mr. S. Kobayashi (Olympus) and for developing and improving the LCM-DIM system. The authors are grateful to Prof. M. Soriaga and Dr. J. Baricuatro (Joint Center for Artificial Photosynthesis California Institute of Technology, Pasadena, United States) for their helpful suggestions and discussion for the paper. This work was supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan under grant No. 20245038 and in part by the New Energy and Industrial Technology Development Organization (NEDO).

 REFERENCES (1)

Gewirth, A. A.; Niece, B. K. Chem. Rev. 1997,

(2) (3)

Itaya, K. Prog. Surf. Sci. 1998, 58, 121. Kolb, D. M. Angew. Chem. Int. Ed. 2001, 40,

1162. (4) Yoshimoto, S.; Itaya, K. Annu. Rev. Anal. Chem. 2013, 6, 213. (5) Itaya, K. Electrochemistry 2015, 83, 670. (6) Więckowski, A. Interfacial Electrochemistry: Theory, Experiment, and Applications; Marcel Dekker, 1999. (7) Alliata, D.; Kötz, R.; Haas, O.; Siegenthaler, H. Langmuir 1999, 15, 8483. (8) Inaba, M.; Kawatate, Y.; Funabiki, A.; Jeong, S.-K.; Abe, T.; Ogumi, Z. Electrochim. Acta 1999, 45, 99. (9) Campana, F. P.; Kötz, R.; Vetter, J.; Novák, P.; Siegenthaler, H. Electrochem. Commun. 2005, 7, 107. (10) Dora, S. K.; Bai, Y.; Elbahri, M.; Kunz, R.; Adelung, R.; Magnussen, O. J. Electrochem. Soc. 2008, 155, D666. (11) Dora, S. K.; Magnussen, O. J. Electrochem. Soc. 2008, 155, F132. (12) Benavente, E.; Santa Ana, M. A.; Mendizábal, F.; González, G. Coord. Chem. Rev. 2002, 224, 87. (13) Stephenson, T.; Li, Z.; Olsen, B.; Mitlin, D. Energy Environ. Sci. 2014, 7, 209. (14) Fang, X.; Hua, C.; Guo, X.; Hu, Y.; Wang, Z.; Gao, X.; Wu, F.; Wang, J.; Chen, L. Electrochim. Acta 2012, 81, 155. (15) Lacey, S. D.; Wan, J.; Cresce, A. v. W.; Russell, S. M.; Dai, J.; Bao, W.; Xu, K.; Hu, L. Nano Lett. 2015, 15, 1018. (16) Xiong, F.; Wang, H.; Liu, X.; Sun, J.; Brongersma, M.; Pop, E.; Cui, Y. Nano Lett. 2015, 15, 6777. (17) Wen, R.; Rahn, B.; Magnussen, O. M. Angew. Chem. Int. Ed. 2015, 54, 6062. (18) Magnussen, O. M.; Polewska, W.; Zitzler, L.; Behm, R. J. Faraday Discuss. 2002, 121, 43. (19) Andersen, J. E. T.; Bech-Nielsen, G.; Møller, P.; Reeve, J. C. J. Appl. Electrochem. 1996, 26, 161. (20) Divisek, J.; Steffen, B.; Stimming, U.; Schmickler, W. J. Electroanal. Chem. 1997, 440, 169. (21) Van Driessche, A. E. S.; Otálora, F.; Sazaki, G.; Sleutel, M.; Tsukamoto, K.; Gavira, J. A. Cryst. Growth Des. 2008, 8, 4316. (22) Wen, R.; Lahiri, A.; Azhagurajan, M.; Kobayashi, S.-i.; Itaya, K. J. Am. Chem. Soc. 2010, 132, 13657. (23) Azhagurajan, M.; Wen, R.; Lahiri, A.; Kim, Y. G.; Itoh, T.; Itaya, K. J. Electrochem. Soc. 2013, 160, D361. (24) Wen, R.; Lahiri, A.; Alagurajan, M.; Kuzume, A.; Kobayashi, S.-i.; Itaya, K. J. Electroanal. Chem. 2010, 649, 257.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(25) Kobayashi, S.-I.; Kim, Y.-G.; Wen, R.; Yasuda, K.; Fukidome, H.; Suwa, T.; Kuroda, R.; Li, X.; Teramoto, A.; Ohmi, T.; Itaya, K. Electrochem. Solid-State Lett. 2011, 14, H351. (26) Azhagurajan, M.; Wen, R.; Kim, Y. G.; Itoh, T.; Sashikata, K.; Itaya, K. Surf. Sci. 2015, 631, 57. (27) Sazaki, G.; Matsui, T.; Tsukamoto, K.; Usami, N.; Ujihara, T.; Fujiwara, K.; Nakajima, K. J. Cryst. Growth. 2004, 262, 536. (28) Sugimoto, T.; Atsumi, Y.; Kikuta, M.; Ishiko, E.; Kono, M.; Ishikawa, M. J. Power Sources 2009, 189, 802. (29) David, L.; Bhandavat, R.; Barrera, U.; Singh, G. Sci. Rep. 2015, 5, 9792. (30) Wagemaker, M.; van de Krol, R.; Kentgens, A. P. M.; van Well, A. A.; Mulder, F. M. J. Am. Chem. Soc. 2001, 123, 11454. (31) Acrivos, J. V.; Liang, W. Y.; Wilson, J. A.; Yoffe, A. D. J. Phys. C: Solid State Phys. 1971, 4, L18.

Page 8 of 9

(32) Papageorgopoulos, C. A.; Jaegermann, W. Surf. Sci. 1995, 338, 83. (33) Wang, Y.; Ou, J. Z.; Chrimes, A. F.; Carey, B. J.; Daeneke, T.; Alsaif, M. M. Y. A.; Mortazavi, M.; Zhuiykov, S.; Medhekar, N.; Bhaskaran, M.; Friend, J. R.; Strano, M. S.; Kalantar-Zadeh, K. Nano Lett. 2015, 15, 883. (34) Adelung, R.; Ernst, F.; Scott, A.; Tabib‐Azar, M.; Kipp, L.; Skibowski, M.; Hollensteiner, S.; Spiecker, E.; Jäger, W.; Gunst, S. Adv. Mater. 2002, 14, 1056. (35) Adelung, R.; Hartung, W.; Ernst, F. Acta Mater. 2002, 50, 4925. (36) Spiecker, E.; Schmid, A. K.; Minor, A. M.; Dahmen, U.; Hollensteiner, S.; Jäger, W. Phys. Rev. Lett. 2006, 96, 086401.


Page 9 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


84x47mm (150 x 150 DPI)


In Situ Visualization of Lithium Ion Intercalation into ... - ACS Publications

Recommend Documents